Cognitive impairments in schizophrenia (SZ) have been linked to dysfunction of inhibitory prefrontal cortex (PFC) circuitry, including parvalbumin (PV) and somatostatin (SST) neurons. We recently found that disturbances in PFC PV&SST neurons are most prominent in a subset (~50%) of SZ subjects identified as a low GABA marker (LGM) molecular phenotype. The etiopathogenesis of PV&SST neuron dysfunction in the LGM SZ subjects may be influenced by genetic liabilities and/or environmental insults that affect neuronal ontogeny. In humans, embryonic PV&SST neurons express developmental regulators (e.g., Lhx6) and cytokine receptors that regulate their specification and migration. We recently found deficits in PFC Lhx6 mRNA levels that were most prominent in the LGM SZ subjects, suggesting that low Lhx6 levels may impair PV&SST neuron development in SZ. In addition, disturbances in the fetal environment, such as elevated cytokine levels due to maternal immune activation (MIA), increase risk for SZ and lower cortical PV levels. These data suggest that prenatal exposure to altered cytokine levels due to MIA may disrupt the development of cytokine receptor-expressing neurons. Since loss of Lhx6 induces deficits in cytokine receptors and MIA lowers Lhx6 levels, the combination of Lhx6 deficits and MIA may severely disrupt PV&SST neuron development. Therefore, we hypothesize that disturbances in PV&SST neurons in SZ subjects with the LGM phenotype reflect the long-lasting consequences of prenatal insults that are fetal (i.e. deficits in developmental regulators such as Lhx6) and/or maternal (i.e. immune activation) in origin. Testing this central hypothesis requires a translational, cross-species approach. In Aim 1 we will use tissue and cellular measures of mRNA levels of developmental factors and cytokine receptors critical for PFC PV&SST neuron ontogeny and other immune markers in the PFC of SZ and healthy subjects. We hypothesize that SZ subjects with the LGM phenotype show a pattern of low mRNA levels for developmental factors and high mRNA levels for immune markers relative to other SZ and healthy subjects. SZ and bipolar disorder (BP) share features including genetic risk, psychosis, cognitive impairments, and low PFC PV and GAD67 mRNA levels. In Aim 2 we will investigate whether a shared pathogenetic mechanism may disrupt PV&SST neuron ontogeny by conducting mRNA studies similar to Aim 1 in BP subjects. We hypothesize that the LGM phenotype and deficits in developmental factors are also present in a subset of BP subjects, but at a lower frequency than SZ. Finally, in Aim 3 we will investigate a potential pathogenetic mechanism that may lead to the LGM phenotype by administering poly I:C which induces cytokine response to pregnant wild-type mice (Lhx6+/- male progenitors) and conducting studies of PFC PV&SST neuron developmental regulators and neurophysiology measures in offspring. We hypothesize that deficits in Lhx6 or MIA independently, and their interaction more severely, leads to deficits in adult PV&SST neurons akin to those seen in the LGM phenotype.